Review: Mehler, Sharoff and Santini (eds, 2011) Genres on the Web: Computational Models and Empirical Studies. New York: Springer

Abstract

Listen

Translate article

Genres on the Web is an edited volume consisting of fifteen studies which delve into a wide range of topics. These include the design and construction of web corpora, user-based and automatic genre identification, and the quantitative and qualitative analysis of web genres based on a variety of structural and linguistic features. This book is an invaluable resource for corpus and computational linguists, as well as those who specialise in areas such as natural language processing, information retrieval and web data mining. Although the book contains studies from several disciplines and dozens of sub-disciplines, this diverse body of scholarly research unites to: (a) recount major accomplishments in web genre identification and description; (b) report on the methods and results of state-of-the-art web genre research; and (c) give explicit direction for future inquiry in this burgeoning field of study. The first part of this book, written by the three volume editors, Mehler, Sharoff and Santini, situates the study of web genres historically and contextually. Major challenges for web genre researchers are introduced in this section, and questions regarding the feasibility of solving specific problems are discussed. After outlining many prominent approaches to web genres, the editors give an overview of the book, which is divided into the following six parts: Introduction, Identifying the Sources of Web Genres, Automatic Web Genre Identification, Structure-oriented Models of Web Genres, Case Studies of Web Genres and Prospect. The second part of the book deals primarily with the issues of identifying the defining characteristics of web genres and classifying those genres according to quantitative and qualitative user-based research. Karlgren begins by proposing that a genre is made up of texts which share a two-way flow of expectations between author and reader (p. 34). He supports this claim by introducing user-based research as a challenging, yet fruitful starting point for genre identification and classification. Next, the results of four consecutive user-based studies are reported. In addition to identifying